Modelling the Influence of Soil Properties on Crop Yields Using a Non-Linear NFIR Model and Laboratory Data

This paper introduces a new non-linear correlation analysis method based on finite impulse response (NFIR) model to study and quantify the effects of ten soil properties crop yield. Two versions NFIR were implemented: NFIR-LN, accounting for both linear variability in system, NFIR-L, only. The performance models was compared with random forest (RF) model, predict oilseed rape (2013) wheat (2014) yields one field at Premslin, Germany. used as system inputs, whereas yield output. Results demonstrated that individual total contribution varied throughout different cropping seasons, weather conditions, crops. Both NFIR-LN RF outperformed NFIR-L explained up 55.62% 50.66% variation years 2013 2014, respectively. performed equally during prediction, although provided more consistent results through two seasons. Higher phosphorus potassium contributions calculated suggesting this outperforms model.